Defining Parameters

Parameters are the typed inputs of a procedure. Laser Setup defines reusable parameters once in parameters.yaml and references them from procedures, so units, limits and descriptions stay consistent everywhere.

PyMeasure parameter types

Parameters are PyMeasure objects. The types you'll use:

Type	For	Example
FloatParameter	real numbers (with units)	vds, laser_v
IntegerParameter	integers	chip_number, N_avg
BooleanParameter	on/off toggles	laser_toggle
ListParameter	a dropdown of choices	chip_group, laser_wl
Parameter	free text / expressions	info, vg ("DP + 0 V")
Metadata	computed values saved but not entered	start_time

The parameters.yaml schema

Parameters are grouped under categories (Chip, Laser, Instrument, Control). Each entry uses a _target_ to select the PyMeasure type — the file defines YAML anchors at the top so you can write *FloatParameter instead of the full path:

_types:
  - &Parameter pymeasure.experiment.Parameter
  - &BooleanParameter pymeasure.experiment.BooleanParameter
  - &IntegerParameter pymeasure.experiment.IntegerParameter
  - &FloatParameter pymeasure.experiment.FloatParameter
  - &ListParameter pymeasure.experiment.ListParameter
  - &Metadata pymeasure.experiment.Metadata

Control:
  vds:
    _target_: *FloatParameter
    default: 0.075
    name: VDS
    description: Drain-Source voltage
    units: V
    minimum: -210.
    maximum: 210.
    decimals: 10

  vg_toggle:
    _target_: *BooleanParameter
    default: true
    name: VG toggle

  vg:
    _target_: *FloatParameter
    default: 0.
    name: VG
    units: V
    minimum: -100.
    maximum: 100.
    group_by: vg_toggle      # only shown when vg_toggle is on

Supported keys

Key	Meaning
_target_	The parameter class (use the anchors).
default	Default value.
name	Label shown in the GUI and written to the file.
description	Tooltip/help text.
units	Unit string (appended to the value).
minimum / maximum	Numeric bounds.
decimals	Display precision.
choices	Options for a ListParameter.
group_by	Show/hide based on another (boolean) parameter.
fget	For Metadata: attribute/function providing the value.

Grouping with group_by

group_by ties a parameter's visibility to a boolean. Two forms:

laser_wl:
  group_by: laser_toggle               # simple: show when laser_toggle is true

step_time:
  group_by: show_more                  # advanced inputs hidden until "Show more"

Procedures also use group_by: {} in procedures.yaml to clear a grouping inherited from the shared definition.

Using parameters in a procedure

laser_setup/procedures/utils.py exposes the instantiated parameters as Parameters, organized by category. Assign them as class attributes:

from .utils import Parameters

class It(ChipProcedure):
    vds = Parameters.Control.vds
    laser_v = Parameters.Laser.laser_v
    NPLC = Parameters.Instrument.NPLC

Why a deep copy?

Parameters is a DeepCopyDictConfig, so each access returns a fresh parameter object. That prevents two procedures from accidentally sharing (and mutating) the same parameter instance.

You can equally define parameters inline (as in the Creating a Procedure example) — handy for one-off inputs that don't belong in the shared file.

Overriding parameter defaults per procedure

procedures.yaml overrides defaults without code. Under a procedure's parameters: key, set any parameter attribute:

It:
  parameters:
    procedure_version: {value: 2.2.0}
    laser_T: {value: 120., group_by: {}}
    vg: {value: "DP + 0. V"}

value sets the default; other keys (units, minimum, group_by, choices, …) patch the parameter object. This is applied by BaseProcedure.configure_class via the @configurable machinery.

INPUTS, EXCLUDE, DATA_COLUMNS, SEQUENCER_INPUTS

These class lists tie parameters to behavior:

• INPUTS — which parameters appear as inputs, in order. Build on the base: INPUTS = ChipProcedure.INPUTS + ['vds', 'vg', ...].
• EXCLUDE — parameters omitted from the saved file header (e.g. transient toggles).
• DATA_COLUMNS — output column names; the first two are the default plot axes. You can append an instrument's columns: DATA_COLUMNS = ['t (s)', 'I (A)'] + PT100SerialSensor.DATA_COLUMNS.
• SEQUENCER_INPUTS — parameters a sequence may sweep.

See the Procedure catalog reference for how the built-ins combine these.